Abstract:During the highspeed magnetic flux leakage testing, the effective magnetization time decreases with the testing speed increase. It may lead to the nonmagnetic saturation of the measured component. The magnetization effectiveness could be affected. To solve this problem, the square wave excitation is used to simulate the external magnetic field transients. The internal magnetic field response model of the steel pipe under the transient magnetic field is formulated. The formulation process and influencing factors of the internal saturation field of the steel pipe are studied. The features of the defect leakage magnetic field during highspeed magnetic flux leakage detection are analyzed. By using finite element analysis, the influences of magnetic field intensity and steel pipe material on magnetization lag time and defect detection are calculated. The highspeed magnetic flux leakage testing platform is designed. Experimental research on the steel pipe defects under different operating speeds and external magnetic field strengths are investigated. Results show that the central magnetic field of the inner wall of the steel pipe obviously lags behind the external magnetic field when the external magnetic field transients. The internal saturation field of the steel pipe is related to the magnetic field strength and the material conductivity. When the intensity of the external magnetic field is increased, the saturation field can be quickly established. Meanwhile, the defect detection effect and magnetic flux leakage detection speed can be improved. There is a good consistence between experiment results and theoretical analysis.

Abstract:In order to achieve accurate measurement of stress information, a new method of existing stress measurement based on inverse magnetostriction for nonexcited steel strand, which is the key loaded component of the prestressed structure, is proposed. Firstly, through establishing the theoretical model of magnetoelectric coupling, the internal relationship of the external load and the permeability of ferromagnetic materials versus the selfinductance coefficient of the coil system is analyzed. Secondly, based on the COMSOL finite element software, the inductance variation of the selfinductance coil system is simulated quantitatively when the stress of the steel strand changes. Finally, the test and measurement platform was built to carry out the test and verification. The results show that the inductance value obtained based on the proposed measurement method has high repeatability; under the condition that the excitation current is constant, the relationships of the inductance versus load for 15 groups of test data all exhibit a significant functional relationship and the goodness of fit R2 is greater than 099. The detection sensitivity increases with the increasing of the effective magnetic circuit length of the selfinductance coil. The error between the fitting value and the measured value is quite small, which meets the engineering requirement and further verifies the rationality of the proposed method.

Abstract:To detect the damage of inner and outer walls of metal plate members, a multichannel lowfrequency electromagnetic sensor designed by means of numerical simulation and experimental optimization. In this study, a threedimensional finite element simulation model of lowfrequency electromagnetic detection is formulated. In this way, the defect detection capability of circular coil placed tangentially over metal plate conductor can be studied. The comparative test is conducted to optimize the number of turns of the excitation coil, the inner diameter and the height of the detection coil, and other parameters, which affect the sensitivity of the sensor. The sensitivity and practicability of the sensor are evaluated by detecting artificial defects contained in the metal plate. Experimental results show that the multichannel lowfrequency electromagnetic sensor can effectively detect the defect of 304 stainless steel with 12 mm buried depth of 96 mm and 20# carbon steel with 12 mm buried depth of 8 mm.

Abstract:Aiming at the problems that great noise effect exists and the inversion ability for multiple magnetic source parameters is poor in traditional magnetic source parameter inversion method, a multimagnetic source parameter inversion method based on magnetic sensor array is proposed. Firstly, the magnetic tensor data measured using the magnetic sensor array are converted into the normalized magnetic source strength (NSS), and the upward continuation of NSS is conducted to reduce the effect of the high frequency noise in the normalized magnetic source strength data. Then, the horizontal position of the magnetic source is estimated using an improved tilt angle based on the normalized magnetic source strength, wherein the maximum value of the positive extremum values in the calculation region is the horizontal position of the magnetic source. Finally, according to the Euler homogeneity equation of normalized magnetic source strength, an improved magnetic source depth estimation method is proposed to realize the single point estimation of magnetic source depth and magnetic moment. The simulation and experiment results show that the proposed method is not affected by the geomagnetic background magnetic field compared with the Nara inversion method. The horizontal location error of multitarget magnetic sources is not greater than 008 m, the depth relative error is not greater than 2476%, and this method also has a strong antinoise ability.

Abstract:The focus thermal drift of micronano CT ray source is one of the important factors affecting image sharpness. The influence of focus drift of the ray source on image sharpness is analyzed through theoretical and simulation experiments. With the actual micronano CT system, experiment discovers that the focus drift is mainly slow thermal drift and the drift is positively correlated with the power of the Xray source, meanwhile the focus drift has certain randomness. On this basis, a focus drift correction method is proposed based on projection image feature matching. Firstly, a small amount of reference projection is rapidly acquired after the actual CT scanning, and the focus drift amount at specific angle is obtained according to result of adaptive feature matching between the actual projection and reference projection. Secondly, all the focus drift amounts in the CT scanning process are obtained with spline interpolation. Finally, the actual projection data are corrected according to the amount of focus drift, and through reconstruction the corrected image is obtained. Experiments show that the proposed method has high positioning accuracy, greatly reduces image distortion and improves image sharpness by nearly 10%.

Abstract:Taking the geometric size of the mobile phone tail plug parts as detection object, a precise measurement method and system of small size, irregular shape parts based on machine vision is proposed. The method includes image acquisition, image enhancement, image registration, edge detection, target line extraction, camera calibration and computational measurement. Aiming at the problems that the traditional scaleinvariant feature transform(SIFT) matching algorithm completely ignores the geometric relationship among different feature points and is prone to more mismatches when searching for matching feature points in the workpiece images with smooth gray change, an improved image registration algorithm is proposed, the contour matching is introduced to acquire image geometric information and constrain the SIFT feature point matching, and the random sample consensus(RANSAC) algorithm is used to remove the influence of noise point pair and precisely estimate the parameters of geometric transformation array. Aiming at the facts that the existing Hough transform fitting line algorithm can easily form pseudopeaks in Hough space for nonlinear edges and affect edge detection accuracy, a new strategy of spatial voting weight allocation in Hough space is designed to suppress the pseudopeak generation. The experiment results show that compared with traditional method, the proposed method improves the accuracy of feature point matching by 12% and the accuracy of line detection by 22%, and the measurement accuracy of the proposed system reaches 0015 mm。

Abstract:In this paper, an arrangement structure description method of multiclass objects based on polar coordinate feature matrix (PCFM) is proposed. The polar coordinate feature matrix consists of polar radius matrix (PRM) and Polar Angle Matrix (PAM), which describes the distance and angle information contained in the arrangement structure of multiclass objects. This method was applied in the application of vehicle fuse box detection and achieved obvious effects. Aiming at the vehicle fuse box detection, the method firstly uses a charge coupled device (CCD) industrial camera to acquire the images of the vehicle fuse boxes, employs histogram of oriented gradient (HOG) feature to characterize the ID codes of the vehicle fuse chips and combines SVMs to achieve the localization and recognition of the fuse chips in the images. According to the localization and recognition results, the PCFM is calculated to describe the arrangement structure of vehicle fuse chips in vehicle fuse box. Finally, the PCFM similarity is taken as the judgment criterion to realize the detection of the vehicle fuse box. The experiment proves that using PCFM similarity to detect the vehicle fuse box, the detection accuracy reaches 976%.

Abstract:For rolling bearing life stage identification, a small number of samples cannot be effectively identified due to the limited sample imbalance under different working conditions. To solve this problem, a multiclassifier integration of the weighted and balanced distribution adaptation method is proposed. Firstly, the training set of multiple samples in source domain is obtained by random sampling, and different initial weights are given to the samples while predicting false labels in target domain. In this way, a few samples can be trained adequately. Then, the classifiers of sample sets in the source domain are trained in the reproducing kernel Hilbert space, and the pseudo labels are optimized. Meanwhile, the weight matrix is updated iteratively. Finally, the strategy of multiclassifier ensemble is achieved. The appropriate base classifier is integrated into a strong classifier to obtain final recognition results. Combining with Fscore evaluation criteria, macroaverage and microaverage evaluation indexes are used to evaluate multiclassification tasks, which can avoid misleading recognition results by accuracy. Experiments on two data sets of rolling bearing life stages verify that the proposed method is feasible and effective.

Abstract:The split pins of highspeed railway catenary positioning tube are easy to be loosed in the longterm running vibration of the train. However, the number of loose samples is scarce. To solve these problems, this study proposes a threelevel cascade architecture expand the defect samples based on deep convolutional generative adversarial network (DCGAN). Then, the convolutional neural network (CNN) is trained to detect split pins defect. Firstly, according to the central point method, the same size image of split pins for training is extracted. Then, DCGAN is used to generate simulated defect samples and a lightweight CNN network is formulated to screen the generated samples. Finally, the extended defect sample set and the positive sample set are utilized to train the detection model on the adjusted VGG16 convolutional neural network. In this way, the defective pins defect state detection can be realized. Experimental results show that the proposed method can achieve 99% accuracy in split pin defect detection of catenary positioning tube.

Abstract:A new time transfer model is proposed to enhance the realtime fault diagnosis performance of rotating machine when the working condition change occurs. Here the source domain is composed of historical data and the target domain is composed of current measurement data. Firstly, the data domains of the model are determined according to the varying working condition rules, and their timedomain feature vectors are extracted to construct the fivedimension spaces. Secondly, the source and target domains are mapped into a twodimension subspace using the maximum variance projection (MVP) and the manifold regularization projection (MRP), respectively. Meanwhile, the minimum mean difference (MMD) criterion is used to minimize the distance between source domain and target domain in twodimension space. Finally, in the projection space, the BP neural network and support vector machine (SVM) classifiers are adopted to build the classification models of the source domain, which are applied in target domain. Also, the diagnostic model is updated through selecting the samples in source domains. Experiments on the gear drivetrain system were conducted, the experiment results prove that the time transfer model can solve the realtime mechanical fault diagnosis problem when the working condition change occurs. Compared with traditional transfer component analysis (TCA) model, the proposed time transfer model can improve the diagnostic performance, the proposed model provides a valuable technical solution for the engineering application of mechanical fault diagnosis.

Abstract:In order to overcome the difficulties that independent selecting key temperature points and performing thermal error modeling destroy their intrinsic relation and reduce the thermal error mode prediction performance, a method of concurrently selecting key temperature points and thermal error modeling under unified framework is proposed. The least squares support vector machine (LSSVM) is used as the basic thermal error model. The selection status of temperature points and the hyperparameters of the thermal error model are regarded as the optimization variables. Furthermore, the binary whale optimization algorithm (BWOA) is used to carry out the optimization. And the cost function is designed by comprehensively considering maximizing the prediction accuracy and minimizing the number of key temperature points. Taking a horizontal machining center as the example, the thermal error experiment was conducted. Using the proposed method, the optimal key temperature points were selected in 10fold crossvalidation mode, the number of key temperature points was reduced from 20 to 3, and the model optimal hyperparameters were simultaneously obtained. Finally, the proposed method was compared and analyzed with the traditional independent method. The comparison results indicate that the thermal error prediction accuracy is improved by 628% at most using the proposed modeling method, which verifies its effectiveness and superiority, and the proposed method also provides a reference for subsequent thermal error compensation implementation.

Abstract:The wind turbine blade usually suffers from fatigue fracture occurred easily under alternating load. So health monitoring of wind turbine blade fatigue damage is very important. A new study method of wind turbine blade fatigue damage is proposed based on thermodynamic entropy. The changing law of the cumulative entropy production of the blade vs. the number of fatigue cycles is analyzed and the threshold point of fatigue fracture occurrence is determined. The results show that the damage energy of the wind turbine blade is also an important factor affecting fatigue damage. There are three stages for the cumulative entropy production of the blade vs. number of fatigue cycles. Study reveals that the threshold point of fatigue damage is the starting point of the third stage of the cumulative entropy production curve. The cumulative entropy production at threshold point and fatigue fracture entropy are independent fixed values, which are not affected by load, frequency and stress ratio. Through calculation, it is found that the ratio of cumulative entropy production to fatigue fracture entropy is 05. Other fatigue experiments were used to verify the threshold point, the result proves that the determined threshold point of fatigue fracture is accurate.

Abstract:Aiming at the deficiency of current feature extraction methods of analog circuit incipient fault diagnosis, the feature extraction method applying deep belief network (DBN) technology is presented. Chaos particle swarm optimization (CPSO) algorithm is employed to optimize the learning rates of the restricted Boltzmann machines in DBN and further improve the feature extraction performance. Compared with other commonly used feature extraction methods, the proposed DBN feature extraction method can extract the deep and essential features of incipient faults. The proposed method also has the features, such as the same high fault aggregation degree and obvious different fault separation capacity. Twostage fouropamp biquad lowpass filter simulation circuit and SallenKey bandpass filter circuit board were used to carried out incipient fault diagnosis experiments, and the obtained fault diagnosis accuracies are 9813% and 100%, respectively.

Abstract:Considering the individual difference and infrequent action in actual operating accessories for the conventional circuit breaker, a realtime remaining mechanical life prediction method based on performance degradation model is proposed. Different from the traditional remaining useful life (RUL) prediction method, this method combines the historical degradation data of the operating accessories with the realtime condition monitoring (CM) data. Firstly, the linear and nonmonotonic degradation process of operating accessories is considered. A performance degradation model is formulated based on Wiener process. Secondly, the historical degradation data are utilized to determine the prior distribution of model parameters by the maximum likelihood estimation method and the onedimensional search method. Then, the Bayesian method is used to update the model parameters iteratively with the realtime CM data. Based on the concept of first hitting time, the RUL prediction model is formulated to realize the realtime prediction. Finally, the proposed method is evaluated by the life data of operating accessories. Experimental results show that the proposed method cannot only realize the realtime remaining mechanical life prediction of the operating accessories, but also has higher prediction accuracy than the methods utilized in other literatures.

Abstract:Aiming at the influence of the inherent hysteresis characteristics of the piezoelectric ceramic actuator on the positioning and control accuracy, a linear equation antihysteresis method based on cutting rate is found. Firstly, the sampling points of the rising and falling trajectories of the hysteresis curve and the cutting rate ratio of the target correction line are calculated, the ratio is the cutting rate coefficient β. The relationship curve between the sampling voltage and corresponding β is fitted segmentally into a linear equation group. Finally, the sampling voltage is inputted into the equation group to obtain the correction voltage, and the driving control curve is made. Experiment results: with the above antihysteresis principle, the maximum hysteresis error of the driver is reduced from 14543% to 1268%, the repeatability error is less than 1497%, and the nonlinear error is less than 4497%. Experiment conclusion: compared with the cumbersome and complex modeling algorithm, the proposed method can realize the correction of the hysteresis curve with the form of proportional amplification and addition operation circuit. The algorithm is only oneorder equation group, which has higher implementation feasibility and operability. The algorithm provides a scientific reference for further improving the positioning and control accuracy of piezoelectric ceramic actuators.

Abstract:In this paper the stability of the differential intersystem biases (DISBs) for the observation values of overlapping frequencies is analyzed based on the observation values of the overlapping frequencies (BDS3 B1C/B2a, GPS L1/L5, Galileo E1/E5a) of BDS3, GPS and Galileo. On this basis, a tightly combined triplesystem realtime kinematic (RTK) positioning method considering DISBs is proposed. The ambiguity resolution (AR) performance and positioning accuracy for the conventional loosely combined model and tightly combined model are compared and analyzed with real test data. The analysis results show that for the baseline composed of the same type of receivers, the pseudorange and carrierphase DISBs among the observation values of the overlapping frequencies for the three systems are very close to zero and thus can be ignored. Compared with those of the conventional loosely combined model where the systems select their individual reference satellites, the model strength, the fixed rate and the fixed accuracy of singleepoch AR for the tightly combined model adopting common reference satellites are all improved to some extent, and the improvement is obvious especially for the obstructed environment with high cutoff elevation angle. In the aspect of positioning accuracy, the tightly combined model also has a certain improvement. In the condition of the cutoff elevation angle of 40°, the positioning accuracies in North/East/Up directions can be improved by 150%, 118% and 194%, respectively.

Abstract:A synchronous on line precision detection method for threedimensional angles (yaw, roll, pitch) is proposed in this study based on the measurement results of standard stepgauge block parameters, which is employed for the fast detection of motion angle errors in linear motion system. The system uses a line laser sensor to achieve the real time acquisition of the stepgauge block parameters. And the mathematical model of the stepgauge block parameters and its threedimensional angles is established. The system measurement stability and repeatability are verified with the calibrated precision turntable. The measured values are compared with the theoretical values, it is determined that within the measurement range of ±30°, the measurement error is less than 1%. The experiment and software calculation combined method was adopted to determine the angle measurement resolution, the result shows that after calibration the angle resolution is 0001°. Finally, the proposed system was used to implement the simultaneous measurement of the three motion angle errors of the zaxis in a threedimensional moving platform, the measurement result verifies that the proposed system possesses the characteristics of simplicity, practicability and high efficiency.

Abstract:It needs to reduce the daily variation (diurnal) in the twoway satellite time and frequency transfer (TWSTFT) and explore the reasons. Using the redundancy of the current AsiaEurope international twoway time comparison network via ABS2A satellite, TWSTFT links of different baseline lengths at different periods are selected. Firstly, the satellite time and ranging equipment (SATRE) TWSTFT results with the softwaredefined radio (SDR) receivers TWSTFT results are compared. Then, with different laboratories in Asia and Europe as a relaying laboratory, the results of SATRE TWSTFT direct and indirect links are compared. Finally, the GPS precise point positioning (PPP) time comparison, which is independent of TWSTFT, is used as a reference to evaluate the performance of the aforementioned results. Experimental results show that the average gain factor of SDR TWSTFT for AsiaEurope and Asia SATRE TWSTFT time deviation (TDEV) is around 175. For the baseline of National Time Service Center (NTSC) and National Institute of Metrology (NIM), the gain factor of TDEV is 122 for the indirect link via PhysikalischTechnische Bundesanstalt (PTB).

Abstract:Emotion recognition is a research hotspot in the field of artificial intelligence. If the humanrobot interaction system can perceive human emotional behavior and express emotion, it will make the interaction between robot and human more natural. Humans acquire emotional information mainly through facial expression, semantic intonation and body language. Taking the NAO robot with high degree of freedom as an application platform, a humanrobot interaction system is designed for facial emotion recognition and body emotion expression. Firstly, the depthwise separable convolution algorithm is introduced to extract and classify features of facial expressions (e.g., angry, fear, sad, happy, surprise and neutral). Results showed that the prediction accuracy of FER2013 facial expression test set could reach 0711 by the trained network model. Secondly, the body movement of NAO robot are designed and classified according to six facial emotions. Finally, the realtime expression of the user′s emotional state by the robot is tested, and the feedback time is within 2 s. The statistical analysis of the prediction results of 10 consecutive frames is carried out.

Abstract:artificial anal sphincter; phase space reconstruction; fast ICA(fast independent component analysis)BP algorithm; rectal perceptual function reconstruction; in vitro experiment; pattern recognition

Abstract:Aiming at the channel selection and classification issue of EEG signals, a motor imagery EEG classification model based on group sparse Bayesian logistic regression (gsBLR) is proposed, which can simultaneously accomplish channel selection and classification. Firstly, spatial filtering and bandpass filtering are performed on the multichannel signals to reduce the influence of volume conduction effect. Secondly, the time domain, frequency domain and timefrequency domain features with discriminant information are extracted for each channel signal, and feature fusion is performed. Finally, the gsBLR method is used for channel selection and classification. The model parameters are automatically estimated from the training data under the Bayesian learning framework, which avoids cumbersome and timeconsuming crossvalidation process. Experiments were carried out on two public BCI competition datasets and selfcollected dataset, and the highest average classification accuracies of 8163%, 8497% and 7647% were achieved, respectively. Compared with other methods, the proposed method achieves better classification accuracy and fewer number of channels. At the same time, the selected channels are more compatible with the neurophysiological background.

Abstract:With the wide application of computers and the development of wearable electronic devices, the combination of humancomputer interaction and wearable devices has become a research hotspot. Gesture recognition technology plays an important role in the field of humancomputer interaction. An intelligent gesture recognition method is proposes based on electrical measurement. The sensor device is used to collect the wrist boundary voltage data. At the same time, the deep voltage neural network is used to classify the collected voltage data, and finally the purpose of gesture recognition is realized. The experiment verifies the feasibility of classifying gestures by electrical measurement data. After adding deep neural network to the gesture recognition system, the correct recognition rate of gestures is over 90%, which proves that the system has better portability, stability and realtime. Sexuality provides new ideas for the design of intelligent gesture recognition systems.

Abstract:Aiming at the problem that wind measurement equipment on the multirotor unmanned aerial vehicle (UAV) platform suffers from the rotorwing turbulent flow interference, based on Bernoulli′s equation, an algorithm is proposed, which utilizes orthogonality type wind pressure vector decomposition to inverse wind speed and wind direction and is suitable to be applied in multirotor UAV. Through designing the layout and structure of the pressure sensing cavity, the impact of the rotorwing turbulent flow is further decreased. The Space claim and Meshing software was utilized to establish the wind measurement model, and the Fluent software was used to change the speed and direction of the airflow in the flow field; the simulation study on the wind measurement method was carried out in different flow fields. During actual measurement, when the wind speed is higher than 2 m/s, the wind measurement error of the proposed orthogonal wind pressure vector decomposition wind measurement method is kept within 10%, which proves that the proposed method could effectively decrease the interference of the rotorwing turbulent flow on wind measurement, and achieve high accuracy wind measurement.

Abstract:It is difficult to separate the overload characteristic signal of projectile penetrating target from complex test signals using the traditional blind source separation (BSS) method. In this study, a new BSS method of penetrating overload signals is proposed, which is not affected by the number of test sensors. This method can also estimate the number of signal sources. Firstly, the singlechannel penetration overload signal is decomposed by the ensemble empirical mode decomposition, and the decomposed intrinsic mode function and the test signal are used to generate multidimensional signals. Secondly, the multidimensional signals are decomposed by the singular value decomposition method. The number of vibration sources is estimated according to the rule of prior Korder singular value dominance. And the maximum crosscorrelation coefficient method is used to determine the best IMF. The test signal and the best IMF are formulated into a multichannel mixed signal. Finally, the multichannel mixed signal is whitened and jointly approximated diagonalzed. The unitary matrix is calculated to obtain the mixed estimation of the original test signal. The acceleration characteristic signal with a correlation of 09747 is obtained by using the method in single channel penetration overload experiment. Compared with the existing methods, this method can effectively separate the characteristic signals of penetration overload. And the adaptive properties of the signal processing process also solves the problem of choosing the filtering frequency of overload signal under different missile target working conditions.

Abstract:Adaptive treadmill is a research hotspot in rehabilitation medicine and ergonomics, and also an important part of virtual reality motion input devices. In this paper, aiming at the problem that the position of the human body on the adaptive treadmill is almost unchanged relative to ground and it is difficult to obtain the speed of the human body through simple position difference, a wide adaptive, unmarked and noncontact walking speed estimation method is proposed. Quaternion calibration, Gaussian filtering and cubic spline interpolation processing are performed on the position data of the human joint points collected with Kinect, the step length correction algorithm is used to calculate the spatiotemporal gait parameters during walking. The speed of the user walking on the adaptive treadmill is estimated based on the gait time and space parameters. The speed estimation value was compared with the actual treadmill speed set on a fixed speed treadmill, the result verifies the effectiveness of the speed estimation algorithm. The proposed speed estimation algorithm can be applied to the development of the control algorithms for adaptive treadmills.

Abstract:Aiming at the problem that the uncertainty of part processing position and feed direction of machine tool causes the change of the tool tip frequency response function (FRF), which leads to the uncertainty of cutting stability lobe diagram and chatterfree processing parameter prediction, a cutting stability prediction and optimization method is proposed combining the support vector regression (SVR) machine and genetic algorithm (GA). This method adopts the hammer impact modal test and spatial coordinate transformation to obtain the tool tip FRFs of different machining positions and feed directions in sample space; then combining the traditional cutting stability prediction method, a SVR prediction model of the limiting cutting depth is established, which takes the displacements of machine tool moving parts, the feed angle, spindle rotation speed, cutting width and the feed rate per tooth as the inputs; the SVR model is taken as the cutting stability constraint to establish the optimization model for the material removal rate (MRR); with the genetic algorithm (GA), the optimal configuration of the displacements of the moving axes, feed angle and cutting parameters is solved. A case study was performed on a certain machining center, and the experiment result shows that the obtained optimal configuration can achieve stable cutting, which verifies the effectiveness and feasibility of the proposed method.

Abstract:Aiming at the accuracy problem caused by the discreteness of composite materials in the dynamics modeling of the wing structure and the influence of the solution speed on the aeroelastic calculation rate of the wing, a method of establishing the dynamic model through combining the modal test and modal method is proposed. In order to improve the solution speed, the wing modal is truncated based on the modal contribution, then the static load numerical calculation and experiment verification for full modal and modal truncation were carried out. Comparing the results of the reduced modal solution and full modal solution, the error of the solution results is only about 025%, and the maximum error of the reduced modal solution is only 60% of the test result. The result shows that the experimentnumerical modeling method can accurately describe the dynamic response of the composite material wing. The modal truncation based on the modal contribution can reduce the model and greatly improve the solution speed without affecting the solution accuracy. The static and dynamic aeroelastic analyses of the wing were carried out. The analysis results show that the aeroelasticity has an unnegligible effect on the dynamic response of the wing.

Abstract:Under complex flight conditions, it is hard to realize the trajectory tracking control of the quadrotor with unmeasurable speed. Considering the existence of unknown external disturbance and uncertain model parameters, the trajectory tracking control method based on an extended state observer is proposed. Firstly, the integral backstepping tracking controller is designed to reduce the steady state error of the system, and the state extended observer is introduced to estimate the unknown speed of the system. Then, the disturbance and the uncertainty of model parameters are estimated in real time and compensated accordingly. Finally, the Lyapunov function is utilized to prove the stability of the control system. Experiments are implemented on the Qball2 platform of Quanser′s quadrotor. Results show that the trajectory tracking controller based on the extended state observer can estimate the unknown speed in the trajectory tracking control process effectively. It can also solve the problem of unknown external disturbance and model parameter uncertainty. The adaptability to the environment can be enhanced. The robustness of the quadrotor to unknown disturbances and the accuracy of trajectory tracking control are improved effectively.